Purpose: We validated a previously reported proteomic signature, associated with treatment outcome, in an independent cohort of patients with non-small cell lung cancer (NSCLC). A novel peptide signature was developed to predict toxicity.
Experimental design: Using automated magnetic C18 bead-assisted serum peptide capture coupled to MALDI-TOF MS, we conducted serum peptide profiling of 50 NCSLC patients participating in a phase II trial of erlotinib and sorafenib. On the obtained peptide mass profiles, we applied a previously described proteomic classification algorithm. Additionally, associations between observed side effects and peptide profiles were investigated.
Results: Application of the previously acquired algorithm successfully classified the new cohort of patients in groups significantly associated with the outcome. The "poor" group exhibited shorter median progression-free survival (PFS) and overall survival (OS) of 1.35 and 1.98 months (with p = 0.00677 and p = 0.00002, respectively) while the "good" group had significantly longer PFS and OS (10.63 and 14.4 months with p = 0.00142 and p = 0.00002, respectively), compared to average OS and PFS. Two specific peptides were detected in the sera of all patients that developed severe toxicity.
Conclusions and clinical relevance: Our results provide an algorithm that, following prospective validation in larger cohorts, could assist treatment selection of patients with NSCLC in the first line setting.
Keywords: Cancer; EGFR prediction; Response; Serum proteomics; Toxicity.
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.